![\"image\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F14946445%2Fa55fd122f77cf7f693682b654ac2cb37%2F1.PNG?generation=1687106893202884&alt=media\")
I sadly run out all my submissions today because of this error. I added an exception statement inside a loop in case there is any error, it will be replaced by empty list for both axes and empty string for chart type. Could anyone please take a look on the code below and also the image of the csv file generated from the test dataset (5 images) and tell me what I did wrong? I highly appreciate your time.
\ndata = []\n\n image, label image_pred_lbls:\n :\n \n \n \n data.append({\n : name[:-]+,\n : x_values,\n : mapping[lbl]\n })\n\n data.append({\n : name[:-]+,\n : y_values,\n : mapping[lbl]\n })\n Exception e: \n traceback.print_exc()\n data.append({\n : name[:-]+,\n : [], \n : \n })\n\n data.append({\n : name[:-]+,\n : [],\n : \n })\n\n\n\npredictions = pd.DataFrame(data).set_index()\n\ndata = []\n\n index, row predictions.iterrows():\n data.append({\n : index,\n : .join((, row[])),\n : row[]\n })\n\ndf = pd.DataFrame(data)\ndf.to_csv(, index=)\n
I sadly run out all my submissions today because of this error. I added an exception statement inside a loop in case there is any error, it will be replaced by empty list for both axes and empty string for chart type. Could anyone please take a look on the code below and also the image of the csv file generated from the test dataset (5 images) and tell me what I did wrong? I highly appreciate your time.
\ndata = []\n\n image, label image_pred_lbls:\n :\n \n \n \n data.append({\n : name[:-]+,\n : x_values,\n : mapping[lbl]\n })\n\n data.append({\n : name[:-]+,\n : y_values,\n : mapping[lbl]\n })\n Exception e: \n traceback.print_exc()\n data.append({\n : name[:-]+,\n : [], \n : \n })\n\n data.append({\n : name[:-]+,\n : [],\n : \n })\n\n\n\npredictions = pd.DataFrame(data).set_index()\n\ndata = []\n\n index, row predictions.iterrows():\n data.append({\n : index,\n : .join((, row[])),\n : row[]\n })\n\ndf = pd.DataFrame(data)\ndf.to_csv(, index=)\n
AFAIK, your lists can not be empty. Replace those with [0,0]
", "rawMarkdown": "AFAIK, your lists can not be empty. Replace those with [0,0]", "votes": null }, { "id": "2308498", "postDate": "06/19/2023 02:58:54", "content": "Instead of using an empty list for both axes in the exception statement, it might be better to use [0,0] for the y-axis and [a,b] for the x-axis. This is because even if you have a single prediction [1], using an empty list will return NaN while scoring.
", "rawMarkdown": "Instead of using an empty list for both axes in the exception statement, it might be better to use [0,0] for the y-axis and [a,b] for the x-axis. This is because even if you have a single prediction [1], using an empty list will return NaN while scoring.", "votes": null }, { "id": "2308582", "postDate": "06/19/2023 04:58:56", "content": "Based on the given values of x and y lists, I should set the chart-type to 'line' since x is ['a','b'] which is string and y is [0,0], right?
", "rawMarkdown": "Based on the given values of x and y lists, I should set the chart-type to 'line' since x is ['a','b'] which is string and y is [0,0], right?", "votes": null }, { "id": "2308587", "postDate": "06/19/2023 05:02:22", "content": "yes, that would work
", "rawMarkdown": "yes, that would work", "votes": null }, { "id": "2308589", "postDate": "06/19/2023 05:05:16", "content": "Thank you so much @aravind36 I wish you the best in this competition.
", "rawMarkdown": "Thank you so much @aravind36 I wish you the best in this competition.", "votes": null }, { "id": "2308672", "postDate": "06/19/2023 06:14:37", "content": "Thank you @mayukh18 , I will replace them with zeros.
", "rawMarkdown": "Thank you @mayukh18 , I will replace them with zeros.", "votes": null }, { "id": "2308678", "postDate": "06/19/2023 06:24:03", "content": "@mayukh18 & @aravind36 I have a follow-up question, please. Imagine your loop goes without any error but your algorithm returns an empty list of x-ticks because it could not read the x values or any other reason. What values should I fill the x list? Do you think I should fill it based on the chart type, e.g., if scatter, then x_list = [0,0], if dot, x_list=['0','0']. Or it does not matter because at the end it will be stored in csv file and everything will be string. What do you think?
", "rawMarkdown": "mayukh18 & @aravind36 I have a follow-up question, please. Imagine your loop goes without any error but your algorithm returns an empty list of x-ticks because it could not read the x values or any other reason. What values should I fill the x list? Do you think I should fill it based on the chart type, e.g., if scatter, then x_list = [0,0], if dot, x_list=['0','0']. Or it does not matter because at the end it will be stored in csv file and everything will be string. What do you think?", "votes": null }, { "id": "2308736", "postDate": "06/19/2023 07:10:07", "content": "Everything should ultimately be converted into string so you have to convert the [0,0] to the correct output string format at the end. Check the sample submission files for reference.
", "rawMarkdown": "Everything should ultimately be converted into string so you have to convert the [0,0] to the correct output string format at the end. Check the sample submission files for reference.", "votes": null } ], "comments": [ { "id": 2308407, "author_name": "mayukh18", "author_url": "", "post_date": "06/18/2023 22:51:51", "content": "AFAIK, your lists can not be empty. Replace those with [0,0]
", "votes": null, "replies": [ { "id": 2308672, "author_name": "hamzahabduljalil", "author_url": "", "post_date": "06/19/2023 06:14:37", "content": "Thank you @mayukh18 , I will replace them with zeros.
", "votes": null, "replies": [] }, { "id": 2308678, "author_name": "hamzahabduljalil", "author_url": "", "post_date": "06/19/2023 06:24:03", "content": "@mayukh18 & @aravind36 I have a follow-up question, please. Imagine your loop goes without any error but your algorithm returns an empty list of x-ticks because it could not read the x values or any other reason. What values should I fill the x list? Do you think I should fill it based on the chart type, e.g., if scatter, then x_list = [0,0], if dot, x_list=['0','0']. Or it does not matter because at the end it will be stored in csv file and everything will be string. What do you think?
", "votes": null, "replies": [ { "id": 2308736, "author_name": "mayukh18", "author_url": "", "post_date": "06/19/2023 07:10:07", "content": "Everything should ultimately be converted into string so you have to convert the [0,0] to the correct output string format at the end. Check the sample submission files for reference.
", "votes": null, "replies": [] } ] } ] }, { "id": 2308498, "author_name": "aravind36", "author_url": "", "post_date": "06/19/2023 02:58:54", "content": "Instead of using an empty list for both axes in the exception statement, it might be better to use [0,0] for the y-axis and [a,b] for the x-axis. This is because even if you have a single prediction [1], using an empty list will return NaN while scoring.
", "votes": null, "replies": [ { "id": 2308582, "author_name": "hamzahabduljalil", "author_url": "", "post_date": "06/19/2023 04:58:56", "content": "Based on the given values of x and y lists, I should set the chart-type to 'line' since x is ['a','b'] which is string and y is [0,0], right?
", "votes": null, "replies": [ { "id": 2308587, "author_name": "aravind36", "author_url": "", "post_date": "06/19/2023 05:02:22", "content": "yes, that would work
", "votes": null, "replies": [ { "id": 2308589, "author_name": "hamzahabduljalil", "author_url": "", "post_date": "06/19/2023 05:05:16", "content": "Thank you so much @aravind36 I wish you the best in this competition.
", "votes": null, "replies": [] } ] } ] } ] } ], "raw_markdown_by_id": { "2308092": "I sadly run out all my submissions today because of this error. I added an exception statement inside a loop in case there is any error, it will be replaced by empty list for both axes and empty string for chart type. Could anyone please take a look on the code below and also the image of the csv file generated from the test dataset (5 images) and tell me what I did wrong? I highly appreciate your time. \n\n\n```python\ndata = []\n\nfor image, label in image_pred_lbls:\n try:\n #...\n # lines of code\n #....\n data.append({\n 'id': name[:-4]+'_x',\n 'data_series': x_values,\n 'chart_type': mapping[lbl]\n })\n\n data.append({\n 'id': name[:-4]+'_y',\n 'data_series': y_values,\n 'chart_type': mapping[lbl]\n })\n except Exception as e: # <------------- exception\n traceback.print_exc()\n data.append({\n 'id': name[:-4]+'_x',\n 'data_series': [], #<---------- empty list\n 'chart_type': ' ' #<---------- empty string\n })\n\n data.append({\n 'id': name[:-4]+'_y',\n 'data_series': [],\n 'chart_type': ' '\n })\n\n\n#------------------------- after the loop -------------\npredictions = pd.DataFrame(data).set_index('id')\n\ndata = []\n\nfor index, row in predictions.iterrows():\n data.append({\n 'id': index,\n 'data_series': ';'.join(map(str, row['data_series'])),\n 'chart_type': row['chart_type']\n })\n\ndf = pd.DataFrame(data)\ndf.to_csv('submission.csv', index=False)\n```\n\n", "2308407": "AFAIK, your lists can not be empty. Replace those with [0,0]", "2308498": "Instead of using an empty list for both axes in the exception statement, it might be better to use [0,0] for the y-axis and [a,b] for the x-axis. This is because even if you have a single prediction [1], using an empty list will return NaN while scoring.", "2308582": "Based on the given values of x and y lists, I should set the chart-type to 'line' since x is ['a','b'] which is string and y is [0,0], right?", "2308587": "yes, that would work", "2308589": "Thank you so much @aravind36 I wish you the best in this competition.", "2308672": "Thank you @mayukh18 , I will replace them with zeros.", "2308678": "mayukh18 & @aravind36 I have a follow-up question, please. Imagine your loop goes without any error but your algorithm returns an empty list of x-ticks because it could not read the x values or any other reason. What values should I fill the x list? Do you think I should fill it based on the chart type, e.g., if scatter, then x_list = [0,0], if dot, x_list=['0','0']. Or it does not matter because at the end it will be stored in csv file and everything will be string. What do you think?", "2308736": "Everything should ultimately be converted into string so you have to convert the [0,0] to the correct output string format at the end. Check the sample submission files for reference." }, "source": "meta" }